dissenting:
I respectfully dissent. In my view, the issue is, once the evidence established that Mrs. Werlich was alive at the time of removal from the respirator, whether the jury should have been instructed that the government must prove beyond a reasonable doubt that the injury Mrs. Werlich received during the robbery would have caused her to die within a year and a day from the infliction of the wound.
I
In every criminal case the government must prove each element of the offense charged beyond a reasonable doubt. E. g., Estelle v. Williams, 425 U.S. 501, 503, 96 S.Ct. 1691, 48 L.Ed.2d 126 (1976); Egan v. United States, 52 App.D.C. 384, 393, 287 F. 958, 967 (1923). As part of this burden, in a homicide case the government must show that the defendant’s conduct is both the actual cause and the legal cause — or proximate cause — of the result. Actual cause means that the defendant’s conduct in fact was the cause of death. Legal or proximate cause means that, although intervening occurrences may have contributed to the death, the defendant can still, in all fairness, be held criminally responsible for that death. W. LaFave & A. Scott, Jr., Handbook on Criminal Law § 35, at 248 (1972). Thus the two, if proven beyond a reasonable doubt, establish that the defendant’s act was the cause of death.
In response to fact situations where two or more actors bring about one death, courts, of necessity, fashioned rules of law to determine which actor is culpable for homicide. These rules reflect an accommodation made by the courts to differing societal interests in punishing certain wrongful acts. Thus, the assailant can be convicted of homicide even though, after inflicting a blow (fatal or nonfatal), (1) the victim fails to receive or accept medical treatment, e. g., Hopkins v. United States, 4 App.D.C. 430, 438-41 (1894); State v. Donges, 126 Mont. 341, 344-46, 251 P.2d 254, 256-57 (1952); (2) the victim fails to follow physician’s instructions, e. g., Warren v. State, 32 Ala.App. 273, 25 So.2d 51, 53 (1946); (3) the victim, because of his injury, takes his own life, United States v. Hamilton, 182 F.Supp. 548, 550-51 (D.D.C.1960); People v. Lewis, 124 Cal. 551, 57 P. 470, 472-73 (1899); (4) a physician administers negligent treatment, e. g., State v. McClain, 256 Iowa 175, 188-89, 125 N.W.2d 764, 771-72 (1964); Devaughn v. State, 232 Md. 447, 454-56, 194 A.2d 109, 113-14 (1963), cert. denied, 376 U.S. 927, 84 S.Ct. 693, 11 L.Ed.2d 623 (1964); (5) or a different or more skillful treatment might have saved the victim’s life, State v. Cox, 82 Idaho 150, 155, 351 P.2d 472, 475 (1960). See generally Annot., 100 A.L.R.2d 769 (1965).
The underlying rationale of these decisions is that the intentional wrongdoer should bear the risk of the victim’s death because the aforementioned intervening acts are considered foreseeable and natural consequences of his wrongful act. See Annot., 100 A.L.R.2d, supra at 783. Implicit in this determination of culpability is a finding that the initial wrongdoer “may fairly be held responsible for the actual result even though it does differ or happens in a different way from the intended or hazarded result.” LaFave, & Scott, supra § 35, at 248.
A more complex issue bearing on allocation of risk and punishment is presented when two or more intentional actors, not in concert, bring about the death of one victim. For example, A seriously wounds B, and, while B is dying, C shoots and kills B instantaneously. Some courts hold that A cannot be liable for homicide because to so hold would necessitate finding that the victim was killed twice. Id. at 250, citing State v. Scates, 50 N.C. 420 (1858); State v. Wood, 53 Vt. 560 (1881). Other courts hold that both A and C can be culpable for *1288homicide because of the combined effects of the wounds, LaFave & Scott, supra at 250, citing People v. Lewis, supra; Commonwealth v. Costley, 118 Mass. 1 (1875), or because the injury inflicted by the first wrongdoer would have also resulted in death without the intervening act. 40 Am.Jur.2d Homicide § 16, at 307 (1968). Nevertheless, as a general rule voluntary infliction of harm by a second actor usually suffices to break the chain of legal cause and protects the original actor from, while exposing the second actor to, culpability for the homicide. See LaFave & Scott, supra § 35, at 259. In sum, when both the original actor and the intervenor inflict harm intentionally, courts differ as to whether the first actor is culpable for homicide, but all agree that the second is culpable. The latter act is considered to have shortened the life of the victim, albeit one who is already dying, and therefore, that act becomes the legal cause of the victim’s death. See id. at 250.
Likewise, rules have been fashioned to allocate risk and punishment between an intentional wrongful act of one person followed by a grossly negligent act of another. As a general rule, gross negligence by an attending physician, which is the sole cause of death, will shield the first wrongful actor from homicide culpability on the theory that the wound inflicted cannot be the legal or proximate cause of death since death results from an unforeseen risk and is not a natural consequence of the wrongful act. Thus, as to the assailant, the causation element of the crime of homicide cannot be proved. See Annot., 100 A.L.R.2d, supra at § 7; LaFave & Scott, supra at 259. See People v. Calvaresi, 188 Colo. 277, 283, 534 P.2d 316, 319 (1975) (en banc).
The aforementioned rules concern themselves with fact situations involving intervening acts that are negligent, grossly negligent, or intentional and wrongful. We are faced today with a different kind of intervening act, one not contemplated by courts of earlier ages. Those courts could neither envision the artificial prolongation of life on life-support systems nor contemplate the cessation of that artificially prolonged life by doctors confronted with most difficult decisions pertaining to such prolongation. As Professor W. S. Peart has written,
Doctors regularly turn off respirators which are maintaining patients who have a normal heartbeat. They do this because they have arrived at a decision that conscious life is no longer possible with the condition present, whether it be a tumor of the brain or . . .a head injury. Since there is no spontaneous respiration, due to the brain damage, the heart will then stop beating due to lack of oxygen and death follows. This decision is being made every day, irrespective of transplantation. [Letter to the Editor, The Times, June 3, 1969, quoted in Skegg, The Termination of Life-Support Measures and the Law of Murder, 41 Mod.L. Rev. 423, 425 (1978) (emphasis by Skegg).]
These are the acts which form the basis of the so-called “pull the plug” cases. These cases require of our jurisprudence an analysis of the principles underlying our legal precedents in this area of the law. We have already noted that courts are generally interested in allocating the risk of the victim’s death between the several actors who in some way contribute to the death. If we strictly adhere to the settled rules, we would have to pigeonhole this intentional act in order to allocate that risk. We could term the act negligent, relieve the physician of criminal culpability, and hold the defendant liable for homicide while ignoring the fact that the physician intentionally hastened the victim’s death. Alternatively, we could term the act an intentional wrong, hold the physician responsible for death, either solely or in combination with the first intentional actor, and thereby ignore the fact that the physician “is serving a socially better end than the individual risking homicide.” The Law of Homicide: Does it Require a Definition of Death?, 11 Wake Forest L.Rev. 253, 266 (1975) [hereafter Definition of Death ]. Clearly, the traditional concepts are not without limitations. To pull the plug on a victim is a unique act. It is an act intentionally done *1289to shorten life. New rules and concepts must be developed not only to protect physicians from an unrealistic application of the law, but also to protect society from assailants who might escape punishment to the full extent of the law for their wrongful acts because of medical advances.
One way to accomplish this end would be to change the definition of death from a cessation of the victim’s heart and lungs to a cessation of brain functioning.1 In this way the physician, with legal immunity, would be able to use the respirator for other patients with some hope of recovery because his conduct of pulling the plug, occurring after death, would bear no legal consequence, while the wrongdoer’s conduct would expose him to homicide liability. Definition of Death, supra at 262.2
Another theory suggests terming the physician’s act as proper medical treatment necessitated by the injury. The act of removing a patient from a respirator would then be considered a natural consequence of the injury and would play no part in legal causation. The physician escapes liability as long as his act is proper, while the assailant can be found culpable for homicide. Skegg, supra at 433-35.3 The majority appears to adopt a version of this theory.
II
In homicide cases, death must ensue from a mortal wound within a year and a day of its infliction. If death does not take place within that period, the law draws a conclusive presumption that the injury is not the cause of death and that death was due to natural causes. If death occurs after a year and a day there can be no prosecution for homicide. 40 Am.Jur.2d Homicide, § 14, at 305; Annot., 60 A.L.R.3d 1323 (1974). This common law rule has not been abrogated by statute and is still in effect in the District of Columbia. Hopkins v. United States, supra at 439, quoting 1 Hale, Pleas of the Crown 428 (1680). See Louisville, Evansville & St. Louis R.R. v. Clarke, 152 U.S. 230, 239, 14 S.Ct. 579, 38 L.Ed. 422 (1894); 1 Wharton’s Criminal Law and Procedure § 191, at 436-37 (1957). Moreover, it is axiomatic that in a criminal case the causation element, as all other elements of the crime, is a matter for the fact finder in all circumstances. Not to allow the jury to determine causation, would be for the court to direct a partial verdict for the government and usurp the role of the jury as fact finder. See Watts v. United States, D.C.App., 362 A.2d 706, 709 (1976) (en banc). To avoid this infringement on the jury’s function as well as to protect the right of the defendant to a jury determination when a dispute arises concerning the cause of death, “the court may and should instruct the jury fully and clearly on the issue.” 41 C.J.S. Homicide § 363, at 142 (1944). The refusal to give the instruction becomes reversible error where the evidence presents a theory of defense — that the injury was not the cause of death since the victim might have lived for more than a year and a day — and the court is apprised of this theory and particularly requested to so instruct the jury. See Levine v. United States, 104 U.S.App.D.C. 281, 282-83, 261 F.2d 747, 748-49 (1958). This rule “applies as well to situations where special facts present an evidentiary theory which if believed defeats the factual theory of the prosecution.” Id. at 282. Further, the defendant is entitled to have the theory of defense instruction even if the evidence that forms the foundation for the theory is “weak, insufficient, inconsistent, or of doubtful credibility.” Tatum v. United States, 88 U.S.App.D.C. *1290386, 391, 190 F.2d 612, 617 (1951), quoting 53 Am.Jur. Trial § 580, at 458 (1945). In essence, as long as the evidence raises the issue, the instruction must be given. Womack v. United States, 119 U.S.App.D.C. 40, 40, 336 F.2d 959, 959 (1964).
Ill
Having set forth the foregoing legal principles, I proceed to my disagreement with the majority opinion. As I understand its holding, it is this: Where an assailant causes injuries to a victim necessitating “heroic measures” for her treatment and a physician decides to pull the plug on the still alive victim within one year of the inflicting of such injuries, to raise a jury issue, the defendant must present evidence 1) that the doctor’s conduct was wrong from a medical standpoint and 2) that the victim would have lived more than 1 year. I respectfully submit that this holding turns the appropriate allocation of the burden of proof on its head. In my view, once evidence is adduced that a conscious, intentional decision was made to terminate life by pulling the plug, it became incumbent that the government be required to prove to the jury, under proper instructions, that this act was not the proximate cause of death and was thus legally irrelevant to appellant’s guilt. Stated another way, the government should have been required to prove, under proper instructions, that if the plug had not been pulled, the victim would have died within a year and a day. Whether the jury, with proper instruction, would have resolved this issue adverse to appellant is not for the court to say. No matter how slight the evidence, the jury must, with proper instruction, find that the government proved its whole case, including causation, beyond a reasonable doubt, before the conviction can stand. Since this issue was not properly framed for the jury under proper instructions, I would reverse.
J.N., Jr.
The failure of the majority to appreciate and properly analyze the alleged error committed by the trial court is more apparent in J.N., Jr.’s case where the trial court precluded any testimony bearing on the fact that the victim was alive at the time of removal from the respirator. The record only reveals, from testimony of the treating neurosurgeon, that
[ajfter seven days, it was determined that the patient’s brain function was irreversibly damaged [and that] it was elected not to pursue additional heroic measures, and therefore, that therapy . . . and other variety of operational procedures were not carried out. And, upon that decision, the patient’s vital signs ceased to function.
The doctor further testified that the mortality rate in this type of injury after more than seven days was 95%.
The defense had reserved the right to recall the neurosurgeon to cross-examine him after a full reading of Mrs. Werlich’s hospital record. From that record, the defense learned that Mrs. Werlich died only after removal from the respirator. It was the defense’s contention that “but for the discontinuance, that for pulling the plug, the victim in this case may well still be living today.”
The court refused to allow the recall of the neurosurgeon for cross-examination on this issue, though admitted, “[I]f the victim had continued to be physically alive for more than a year and a day following the alleged attack there would be no case of murder before the court. There might be some other charges, but that would not be one.”
When the court refused to reconsider this ruling, it stated:
[Assuming for the purposes of determination of the motion to reconsider that [the neurosurgeon] were to be recalled and would testify that heroic measures were discontinued rather than none were employed, the court nevertheless would be faced, if the court accepted that testimony . with speculating as to when and in the face of the fact that is plain before all of us that this victim did in fact die, that she died under circumstances which have been evaluated by medical testimony that would effect 95% *1291of all people in the same way, regardless of age.
******
[T]he court is saying . . . that the introduction of that testimony, even if taken as true, does not revive the court with a basis to speculate that life would have continued for a period beyond which the charge in this case would be decisive than what it is. And the court will not reconsider its determination and will not allow the doctor to be called for that purpose.
It is too easy to suggest that the causation element was proved by the government beyond a reasonable doubt because Mrs. Werlich in fact died several days after the injury. As was previously stated, the nature of the physician’s act requires that the government prove that Mrs. Werlich would have otherwise died from the appellant’s wound within a year and a day from its infliction. The appellant should have been given the opportunity of raising a reasonable doubt, whether he was successful or not, that his assault on Mrs. Werlich was not the legal cause of her death — that the injury inflicted would not have caused her to die within a year and a day, and that it was only the doctor’s intentional act that assured her death within this period that established appellant’s culpability for homicide. To refuse to permit the witness to be recalled to give the proffered testimony constituted an abuse of discretion mandating reversal. See generally Johnson v. United States, D.C.App., 398 A.2d 354 (1979). Since the majority declines to do so, I dissent.
. For further discussion of brain death and its effects upon both doctor and defendant, see Guthrie, Brain Death and Criminal Liability, 15 Crim.L.Bull. 40 (1979).
. As the majority correctly perceives, such an approach would be of no assistance in this case since all concede that the victim was alive under any legally accepted definition of death when the plug was pulled.
.Professor Skegg broadens a theory first expressed by Lord Devlin in Samples of Lawmaking at 95 (1962). Lord Devlin was speaking of proper and necessary treatment to relieve pain and suffering. Professor Skegg includes within this theory withdrawal of a patient in an irreversible comatose state from a respirator. Skegg, supra at 433.